Скачать презентацию AERC International Conference 2016 Democracy-Governance-Corruption Nexus Empirical Evidence Скачать презентацию AERC International Conference 2016 Democracy-Governance-Corruption Nexus Empirical Evidence

80519031394aa2a62b021f05169612ef.ppt

  • Количество слайдов: 29

AERC International Conference 2016 Democracy-Governance-Corruption Nexus: Empirical Evidence from Developing Countries Saranjam Baig, Ph. AERC International Conference 2016 Democracy-Governance-Corruption Nexus: Empirical Evidence from Developing Countries Saranjam Baig, Ph. D Karakoram International University, Gilgit-Baltistan Professor Yi Feng, Ph. D Claremont Graduate University, California

When no one is corrupt, no one will be poor. ” (Benigno Aquino III When no one is corrupt, no one will be poor. ” (Benigno Aquino III President of the Philippines, 2010)

Introduction and Background – Corruption # 1 issue in the world. – People may Introduction and Background – Corruption # 1 issue in the world. – People may not be corrupt, but the systems are! – What’s behind that: • POOR GOVERNANCE – Which Governance? • • I mean what we mean by governance here? C = M+D-A Democracies have lesser Corruption? Implicit assumption that democracies have good governance.

Introduction and Background • Rose-Ankerman (1999, 2004, 2012) Corruption may not be because of Introduction and Background • Rose-Ankerman (1999, 2004, 2012) Corruption may not be because of dishonesty of people but inefficiency. • Corruption Vs Governance? Confusion? • Focus on other factors like democracy • World Bank’s Governance Effectiveness measure – Civil Service Reforms – Quality Policy Design and implementation – A holistic measure

Literature Review • Why to study determinants of Corruption? – When there is a Literature Review • Why to study determinants of Corruption? – When there is a huge academic scholarship on it? – Treisman (2000) “The causes of corruption: a crossnational study” – “Grease the wheels” hypothesis: • Leff (1964), and Huntington (1968) – In the absence of well-functioning institutions – Missing in the current literature: » Importance of Institutions of governance in controlling corruption empirically not tested. » Specific focus on developing countries?

Why to focus on developing countries? – “Corruption in Sweden and Nigeria are just Why to focus on developing countries? – “Corruption in Sweden and Nigeria are just not the same. To use the same for completely different situations can only generate confusion. ” Ocampo (1993) – The persistence and homogeneity of rampant corruption in developing countries suggest that they share common drivers of corruption (Khan, 2006).

Why to focus on developing countries? – Khan (2006)’s analytical framework: • developing countries Why to focus on developing countries? – Khan (2006)’s analytical framework: • developing countries following the policies of democratization, low government interventions, active civil participations, and wage compensations without building state governance capacities will not reduce levels of corruption. – State Capacity? – democracies, in themselves, are partially endowed with the capacities necessary for maintaining government legitimacy (Xin and Rudel 2004) • A measure of revenue extraction through the broad-based taxation of citizens as a proxy measure for the ability of a state to mobilize its citizens, and in turn, to support the state.

Primary Questions • Does democracy is a corruption deterring institution in itself ? • Primary Questions • Does democracy is a corruption deterring institution in itself ? • Does governance effectiveness (policy effectiveness) affect corruption? • Is the effect of democracy on corruption conditional upon policy effectiveness? • Does governance capacity (political capacity) affect corruption? • Is the effect of democracy on corruption conditional upon political capacity? • Whether policy effectiveness and political capacity are substitutes?

Theory • The Principal-Agent Model: – Principal: Voters • Punish the corrupt • Reward Theory • The Principal-Agent Model: – Principal: Voters • Punish the corrupt • Reward the honest – Agents: Politicians - Bureaucrats Nexus – Making the opportunity cost expensive – In most developing countries this relationship is weak – Making the principal-agent model workable? • Electoral reforms a sufficient condition? • Policy reforms a necessary condition?

 • UP(c) = U(c) + ρ(c, ß) δVP (1) • UP(c) = U(c) • UP(c) = U(c) + ρ(c, ß) δVP (1) • UP(c) = U(c) + ρ(c, ß*Ġ) δVP (2) • UO(c) = U(c) + ρ(c, ß*Ġ) δVO (3)

Theory • Aidt et al. (2008) – “While in autocracies replacement often takes place Theory • Aidt et al. (2008) – “While in autocracies replacement often takes place through coups, revolts or revolutions. Rulers are willing, up to a point, to reduce corruption today to avoid replacement and loss of future rents, but only where institutions are of a sufficiently high quality. ” • Hadenius and Teorell (2007) – we consider authoritarian regimes, as without considering them any discussion of corruption in developing countries would be incomplete. As roughly half of the countries in the world, particularly in the developing world, can be classified as authoritarian in one way or another, Hadenius and Teorell (2007) – China? – India?

Data and Empirical Model • Sample size: 94 -98 Developing Countries • OLS, 2 Data and Empirical Model • Sample size: 94 -98 Developing Countries • OLS, 2 SLS, WLS • Aggregated Approach: – You & Khagram (2005) • Panel Data/Fixed Effects ? • Dependent variable: Average 2002 -2010, 2011, 2012, 2013, 2014, Average 2011 -2013 • Independent Variables: Average 2002 -2010, 2010

Main Empirical Model Corruption = β 1 Democracy (Freedom House or Dummy) + β Main Empirical Model Corruption = β 1 Democracy (Freedom House or Dummy) + β 2 Governance Effectiveness + β 3 Democracy × Governance effectiveness + β 4 Trade openness + β 5 Rents from natural resources (%GDP) + β 6 Government revenue (%GDP) + β 7 Logged GDP per capita + β 8 Logged Population + β 9 Ethnic fractionalization(index) + β 10 Religion fractionalization(index) + β 11 Region Dummies + μi

Measuring Corruption • Corruption is done behind closed doors. cannot just ask public officals Measuring Corruption • Corruption is done behind closed doors. cannot just ask public officals how much they earn from bribes. • So, use proxies – Expert judgments – Surveys of citizens or firms – Ground-level studies of behavior (cf. Banerjee)

Examples Single source • ICRG • Freedom House • Polity IV • Gallup World Examples Single source • ICRG • Freedom House • Polity IV • Gallup World Poll • Afrobarometer • Latino-Barometer Composite Index • Transparency International (CPI) • World Bank Governance Indicators (WBGI) • Ibrahim Index of African Governance

World Bank Governance Indicators • Widely used by academics, media, and policymakers • Covers World Bank Governance Indicators • Widely used by academics, media, and policymakers • Covers 6 areas: 1) Voice & accountability 2) Political instability & violence 3) Government effectiveness 4) Regulatory burden 5) Rule of law 6) Control of corruption • Uses surveys from 35 sources to create a mean and confidence interval for each country. 212 countries

Transparency International’s Corruption Perceptions Index • • • About 180 countries ranked annually in Transparency International’s Corruption Perceptions Index • • • About 180 countries ranked annually in terms of perceptions of freedom from corruption (0 -100, higher scores are better). Data are averaged between current year and previous year A composite of 13 different surveys conducted by 11 independent international organizations. Both business people and country analyists/ experts – resident and non-residents are polled.

 • So, TI’s CPI combines data from risk assesment firms, expert ratings, and • So, TI’s CPI combines data from risk assesment firms, expert ratings, and surveys to create a standardized score and st. dev. Includes all countries that have ≥ 3 sources of data available. • Countries are ranked each year (but N of countries differs over time). • Key differences from WGI – at least 3 sources and sources are not weighted. How different are these two measures of corruption?

How they behave empirically All correlations are ≥ 0. 75, How they behave empirically All correlations are ≥ 0. 75,

Key Findings (a) • Democracy: – The effect of democracy on corruption is not Key Findings (a) • Democracy: – The effect of democracy on corruption is not robust – Sensitive to additional control variables – Sensitive to different samples – The linear relationship is not robust – Looses its significance once control for governance effectiveness • Interaction of Democracy and Governance effectiveness – Always significant with the expected negative sign

Key Findings (b) • Democracy: Key Findings (b) • Democracy:

Key Findings (b) • Democracy: Key Findings (b) • Democracy:

Key Findings (c) • Democracy and Governance: Key Findings (c) • Democracy and Governance:

Key Findings (c) • Democracy and Governance: Key Findings (c) • Democracy and Governance:

Key Findings (c) • Democracy and Governance: Key Findings (c) • Democracy and Governance:

Limitations • Causality and endogeneity – Instruments? • Not easy to find! – Interpretations Limitations • Causality and endogeneity – Instruments? • Not easy to find! – Interpretations • Subjective measures • Measurement error of corruption indices

Conclusion and Policy Implications • Electoral reforms alone are not sufficient to mitigate corruption Conclusion and Policy Implications • Electoral reforms alone are not sufficient to mitigate corruption • By just reforming one segment of the governance, governments in the developing countries can drastically reduce corruption---both political and bureaucratic.

Further Research • In this research our primary focus has been remained on governance, Further Research • In this research our primary focus has been remained on governance, democracy, and their interaction. However, if the influence of democracy is conditional upon good governance, there is a reason to believe that the effect of economic policies, such as foreign aid and trade openness, can be conditional on the effectiveness of governance. This is open for further research.